[R] correlation coefficient

[Dimitri Liakhovitski]

Just another opinion about R^2 coming from the field of US Psychology
research and business:
The first and foremost technique taught in Psychology Departments in
subfields where experimental designs are rarely possible (i.e., social
psychology, personality psychology, developmental psychology,
Industrial/Organizational psychology) is multiple regression. You have
some important response variable and a bunch of predictors. The
goodness of fit of the model is assessed by looking at R^2. Hence,
everyone has a feel for it and everyone wants to see it reported.
Further, in business settings, most clients can understand a simple
correlation. So, "correlation squared" is relatively easy to explain.
Most of the corporate clients get nervous if you try to explain even a
simple confusion table. Why? Because they have no internal benchmark
for what's good fit and what's bad fit. With R^2 it's much easier.
Dimitri


On Wed, Apr 29, 2009 at 7:51 AM, Peter Flom
<peterflomconsulting at mindspring.com> wrote:
> Dieter Menne <dieter.menne at menne-biomed.de> wrote
>
>>I noted the "and" was misleading. Read: Good journals like Lancet,
>>New English and many British Journal of XXX really help you to do
>>better.
>
> I am one of the statistical editors for PLoS Medicine, and I try to help
> people do better; often, the people take my advice.  Sometimes, they don't.
>
> I get good support from the editorial people there.
>
> Peter
>
>
>
> Peter L. Flom, PhD
> Statistical Consultant
> www DOT peterflomconsulting DOT com
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>



-- 
Dimitri Liakhovitski
MarketTools, Inc.
Dimitri.Liakhovitski at markettools.com